Presentation at the CCIRN meeting in Xi ’ an August 26, 2007

1. Presentation at the CCIRN meeting in Xi’an August 26, 2007 On the Education And Research Networking Evolution STudy –EARNEST – a foresight study follow-up of SERENATE

2. Three phases of EARNEST March-June 2006: • A preliminary phase identifying topics and people to be involved, with an initial workshop in May July 2006 – August 2007: • Seven parallel sub-studies • Two workshops for NRENs resp. Governments August- October 2007: - Integration and summary, conclusions and recommendations, with a final workshop in September

3. The Seven parallel sub-studies • Researchers’ Requirements • Technical Issues • Campus Issues • Economic Issues • Geographic Issues • Organisation and Governance Issues • Needs of Users in Schools, Healthcare, and the Arts, Humanities and Social Sciences

4. Researchers Requirements • This sub-study has been led by the European Science Foundation (ESF) • 2 surveys have been conducted: - in the first stage, more than 4000 researchers have filled in a questionnaire - in the second stage was a more focused survey was made using interviews The surveys have shown that having access to highspeed networking has had a substantial impact , eg broadened the research areas for scientists in most research fields, making new interdisciplinary studies possible, and that the demand for new services is rising. Scientists also show significant interest in enhanced support and training in the usage of networking

5. Campus issues • This sub-study has been led by EUNIS, the European University Information Services association • It has a number of interesting results, eg that the campus network is no longer experienced as a bottleneck in the network provision – as it was in the SERENATE study • But there is a need for further collaboration between campus IT staff and NRENs, and a need for better training and funding for campus IT staff

6. Technical Issues • The study area focuses on possible technical developments relevant to research and education networking in the next few years • The SERENATE technology study largely focused on lower-layer issues, but currently control techniques are once again attracting attention as optical switching and hybrid networks become ubiquitous • Also middleware is a new topic for study, as authentication and authorization have become important issues, complicated by the fact that users have become increasingly mobile • The next 5 slides will go more in depth with the findings of this sub-study

7. Transmission Findings • New low loss fibre promises longer transmission distances (up to 150 km without amplification), whilst new G.656 standard supports CWDM and additional DWDM channels. • G.655 and G.656 fibre may be obtainable, but most routes will likely be mix of older fibre types. May therefore be difficult to take advantage of recent advances in transmission capabilities. • No obvious path for SDH beyond OC-768 (40 Gbps), and likely to become legacy technology in coming years. • All manufacturers developing 40 and/or 100 Gigabit Ethernet because of cost advantages (an OC-192 port is 10 times that of a 10 GE port). Carrier-grade OAM&P and virtual circuit functionality is currently being added (e.g. PBBTE, CFM) • 100 GE implementations expected by 2010, although may be later. Initially likely to be 4 x 25 Gbps and restricted to short-haul applications. Full serial implementations not expected until 2012. • 40 GE may be interim solution as implementations expected by 2009 (expected to be 40% the cost of OC-768). Supposedly for data centre applications, but some vendors talking about WAN capabilities (80 km before amplification/2000 km before regeneration). • Most manufacturers focusing on 50 GHz spacing for DWDM channels (i.e. ~80 channels per fibre). This has been found to provide good performance trade-off with respect to faster line rates and longer reaches.

8. Control Plane and Routing Findings • Routing scalability becoming problematic (again). • Huge rise in number of hosts, fragmentation of service provider hierarchy, increase in multihoming, and amount of traffic. • Global routing table now >230,000 entries, which generates around 400,000 BGP updates per day. • Concern that growth is starting to outstrip router chipset and memory developments, but more specifically the cost of provisioning these. • Not immediate cause for concern, but IAB/IETF looking for efficiencies. Proposed to split IP addresses into identifiers and locators. [Possible implications for AAA as well] • IPv6 growth has been slow, but new predictions now suggest IPv4 address space could be exhausted in 3-5 years. • IPTV may (finally) provide impetus for IP multicast. • MPLS widely adopted, but GMPLS less popular.

9. Network Virtualisation Findings • Virtualisation concepts starting to be used across all networking layers. • Basic virtualisation already implemented in certain modern routers to enable upgrades and troubleshooting of specific interfaces. • NRENs (e.g. CANARIE, CESNET) pioneered customer-empowered network concept, where multiple virtual networks can be defined and managed by customers over NREN-provisioned infrastructure. • Deployment of UCLP and similar technologies are first step towards full network virtualisation. • GENI and FEDERICA initiatives aim to develop network virtualisation to allow disruptive technologies to be tested over production infrastructure.

10. Operations and Performance Findings • Limited tools for managing Network Layers 0-2, and very expensive. • Management of Layers 0-2 currently labour intensive and relies heavily on documentation. • Core networks likely to continue to be overprovisioned as bandwidth is cheap, although should support QoS information so QoS can be applied on edge networks if necessary. • Increasing availability of dark fibre allows R&E networks to operate hybrid networks, enabling dedicated links to be provisioned for demanding customers using C/DWDM. • Middleboxes such firewalls, NATs, rate shapers and other ‘black box’ solutions are responsible for many network problems. They also encourage workarounds that circumvent what the box is trying to achieve in the first place. More careful use and management needed. • Most end-to-end performance issues are due to problems at customer sites. • 10 Gbps+ network monitoring is currently expensive, but cost should reduce with specialist multi-core appliances. • Recommend extending scope of PERT (possibly integrating with NOCs).

11. Middleware Findings • Identity federations are solution for supporting user access to remote services. • Most NRENs have identity federation or are establishing one. Others should plan to do so within next couple of years. • NRENs are natural candidates for supporting technical organisation within their countries, as well as representing national federations. • User-centric identity management also growing, and abstract identity framework also being worked on. NRENs should monitor developments. • It is recommended that NRENs support multiple trust infrastructures in order to be able to handle different AAIs, although should try to minimise number necessary (e.g. by reusing existing PKIs). • SAML 2.0 likely to be used for exchanging identity data for web-based applications. X.509 certificates still used for other applications. Authorisation decisions supported by schemas such as eduPerson or SCHAC. • No well established standard for communicating identity data to applications. NRENs should be proactive about this (possible TF?)

12. The other sub-studies • The Geographic Issues - here a new index, called REDI, has been developed – throwing new light on the concept of ”digital divide” • The Organisational and Governance Issues – points eg to the importance of the involvement of end-users in NREN development • The Economic Issues points to two major future changes that may affect the economics of research networking – one being the investment in dark-fibre in Europe, the other the development of a more elaborate portfolio of services

13. The other sub-studies continued • The sub-study on other users has interesting observations about the developments in • Schools networking – a number of European NRENs now also serve schools and this gives synergy to the research networking, especially in smaller countries • Healthcare networking – here NREN experiences can benefit the advanced services to the healthcare sector, ie there is a basis for transfer of knowledge to this research related sector • Arts, Humanities and Social Sciences are developing new and interesting applications of highspeed networking

14. Final workshop and wrap-up • We are now in the last phase of the EARNEST study, with the final sub-study reports just about there • The final workshop will take place in Brussels (La Hulpe) September 25-26 • After that, the Summary report will be written with a number of recommendations at several levels – including the campus level, the national level and the European level